1

In the paper "Deep Residual Learning for Image Recognition" the authors make use of, so called, shortcut connections. How do they work and where could I find more information about them?

cerebrou
  • 215
  • 2
  • 6
  • These are links passing the same information ('identity' means that this information is not altered). – Alleo Jan 21 '16 at 03:20
  • Check out this comment here about skip layer connections in the recent MSR article: http://stats.stackexchange.com/questions/56950/neural-network-with-skip-layer-connections/189534#189534 – Indie AI Jan 25 '16 at 14:19

0 Answers0